
	Version PDF

Drug Hypersensitivity Syndrome in a West‐Indian population

European Journal of Dermatology. Volume 13, Number 5, 478-81, September 2003, Clinical report

Summary

Author(s) : Philippe MULLER, Patrick DUBREIL, Antoine MAHÉ, Isabelle LAMAURY, Birgit SALZER, Jacqueline DELOUMEAUX, Michel STROBEL , Dermatology, Infectious Diseases and Public Health Departments, University Hospital, BP 465, 97159 Pointe à Pitre, Guadeloupe, French West‐Indies. .

Summary : Some studies have suggested an ethnic susceptibility to Hypersensitivity Syndrome. We did a 7‐year‐prospective study in Guadeloupe whose population is mainly of African ancestry, and has free access to modern care facilities. Most patients included were Afro‐Caribbeans (26\\28), and females (20\\28). However, ethnic distribution did not reach significant conclusions. Annual incidence rate was estimated at 0.9\\100,000. Medium incubation and duration were 33 and 66 days respectively. Two patients with grade 4 hepatitis died from the syndrome. Two thirds of the patients were given prednisone, which usually alleviated the systemic symptoms, but did not prevent their development (in 5 patients) nor death. Carbamazepine, allopurinol, and minocycline accounted for 2\\3 of the cases. Sixty four percent of the causative prescriptions were judged inappropriate. DHS appeared as the most frequent type of severe systemic drug reaction in this population, and may largely be prevented by rational prescribing.

Keywords : drug hypersensitivity syndrome, epidemiology, carbamazepine, allopurinol, minocycline

ARTICLE

Auteur(s) : Philippe MULLER¹, Patrick DUBREIL¹, Antoine MAHÉ¹, Isabelle LAMAURY², Birgit SALZER¹, Jacqueline DELOUMEAUX³, Michel STROBEL²

¹ Dermatology, ² Infectious Diseases and ³ Public Health Departments, University Hospital, BP 465, 97159 Pointe à Pitre, Guadeloupe, French West‐Indies.

Reprints: P. Muller Fax: (+590) 89 16.15. e‐mail: philippe.mullerchu‐guadeloupe.fr

Article accepted on 17\7\03

Drug Hypersensitivity Syndrome (DHS) ‐‐ also known as "Drug Rash with Eosinophilia and Systemic Syndrome" (DRESS) ‐‐ is a rare, idiosyncratic reaction with a delayed onset, a prolonged course, and a case fatality rate around 10% [1‐3]. Diagnosis is merely clinical and relies on four non specific criteria: a) a documented exposure to a drug; b) fever and rash; c) systemic involvement, notably liver, lymph nodes, kidneys or other organs; d) eosinophilia > 500\mm3, and\or mononucleosis‐like lymphocytosis. DHS may be confused with protracted viral infection, auto‐immune disease, hypereosinophilic syndrome, and lymphomas. It is closely related to the "Drug Induced Pseudo Lymphoma" [3, 4]. It is not specific to a single drug or drug class. Although the number of causative drugs is expanding, only a few of them steadily rank on top of the list, including: aromatic anti‐epileptics such as phenytoin (first reactions reported in 1939), phenobarbital, and carbamazepine (the "anticonvulsivant syndrome") [5], dapsone (1950), and allopurinol [6, 7]. DHS has regained attention after recent reports focusing on minocycline [8], and also on the two antiretroviral drugs nevirapine and abacavir [9]. Pathophysiology is largely unknown [10]. Previous studies have suggested a higher prevalence in some populations, mainly Melanesians or Africans [1, 11‐13].

Patients and methods

In order to assess incidence, occurrence pattern, clinical features, outcome, and causative drugs of DHS, we conducted a prospective study, over a seven‐year‐period (1994‐2000

), in the three units of internal medicine, infectious diseases, and dermatology of the only tertiary hospital of Guadeloupe island (University Hospital, Pointe à Pitre, French West Indies). This overseas territory has a tropical environment, an intermediate economy, an easy access to modern medical facilities, and a population of 440,000 (census 1998), mostly of African ancestry (around 75% being Afro‐Caribbeans), with minorities originating from India and France. Ethnic affiliation was defined by the patients themselves.

The subjects were selected among inpatients referred for or suspected of severe systemic drug reaction. A case was defined as DHS if it met the following: a) all four definition criteria listed above; b) a clinically symptomatic course ≥ 10 days; c) a follow‐up ≥ 3 months; d) no other identifiable cause for the syndrome. Children under 15, AIDS patients, those with non systemic patterns (peptic ulcer caused by anti‐inflammatory drugs, or bleeding due to anticoagulants) were excluded. For each patient, a complete medical history was recorded, including the exhaustive list, chronology, and indications of the drugs taken in the previous 3 months. A complete clinical examination was performed as well as repeated biological tests including: routine blood count and chemistry, blood and urine cultures, antinuclear factors, and serology for CMV, EBV, HIV, HTLV1, syphilis, viral hepatitis, toxoplasmosis. Skin or liver biopsy, or lymph node aspiration, were performed when judged useful for diagnosis or prognosis. Drug imputability was based on a combination of intrinsic (chronology, clinical presentation) and extrinsic (evidence based) criteria according to Begaud et al. which are consensual in France [14, 15]. And finally, the prescription appropriateness was assessed according to "Dictionnaire Vidal", the official reference textbook on drugs [16]. Each prescription was then classified as either "appropriate" if it was in accordance with the approved indications; or "inappropriate" if it was not, or if it has been subject to caution or restriction in the recent literature according to a systematic Medline search covering the years 1995‐2000.

Results

Twenty eight patients were included as DHS: 20 female, 8 male, median age 49 years (range 15‐87). Their clinical and biological characteristics are listed in Table I.

Table I. 28 DHS cases. Clinical and biological data

Clinical and Biological Spectrum	Frequency N (%)	Medium values ± SD (Range)
Delay of onset Duration		33 ± 22 days (8‐90) 66 ± 41 days (14‐275)
Fever and rash	28 (100%)	Fever 39.5 ± 0.7°C (37.5‐40.5)
Facial oedema	20 (71)
Mucous membranes involvement	24 (86)
Lymphadenopathy	21 (75)
Hepatitis	27 (96)	ALAT: 184 ± 180 µ\ml (36‐1300)g γGT: 289 ± 237 u\ml (38‐880)
Renal involvement	13 (46)	Creatinine: 138 ± 55µmol\l (80‐270)
Pulmonary (clinical or imaging)	10 (36)
Lethal outcome	2 (7%)
Raised ESR (> 50 mm\h)	8 (28)	52 ± 37 mm\h (20‐135)
C Reactive Protein (> 30 mg\l)	15 (53)	96 ± 72 mg\l (6‐330)
Leukocytes > 10.0 G\l	20 (71)	16.5 ± 8.0 G\l (4.5‐35)
Eosinophils > 0.5 G\l	27 (96)	3.9 ± 3.6 G\l (0.6‐13)
Lymphocytes > 5.0 G\l	13 (46)	4.6 ± 2.8 G\l (0.8‐13)
Hyperbasophilic lymphocytes > 5%	14 (50)	(0‐32%)

.

Twenty six were Afro‐Caribbeans, 2 Indo‐Caribbeans, and none was Caucasian. Incidence of the syndrome was estimated at 0.9\100000\year. These 28 DHS cases accounted for quasi half (47%) of a total of 59 severe systemic drug reactions referred during the study period including: 15 Lyell\Stevens‐Johnson syndromes, 10 drug‐induced grade 3‐4 hepatitis, 2 drug induced lupus erythematosus, and 4 unclassified. Medium delay of onset (from first drug exposure to first symptom) was 33 days. Noteworthy, all cases had a similar clinical presentation irrespective of the incriminated drug. Fever and rash were constant features. The rash was extensive, pruritic, occasionally recurrent (2 cases), and exhibited 3 different patterns: a) an initial maculo‐papular rash, noted in all cases, and associated with oral mucous membrane involvement (86%) and marked facial oedema (71%); b) erythroderma in 8 cases (28%), either congestive or pustular; c) target lesions or bullae (32%) suggestive of erythema multiforme (EM) (6 cases), or Stevens‐Johnson\Lyell syndrome (3 cases). Skin biopsies were obtained from 15 patients. Histopathologic changes were non‐specific, showing mild dermal vasculitis and epidermal basement membrane vacuolisation, a picture resembling EM. Two specimens however exhibited a lymphoma‐like pattern with a dermal infiltrate of large dystrophic lymphocytes showing some epidermotropism. Lymphadenopathy was present in 21 patients (75%), and also led to the suspicion of lymphoma in 5 cases: node aspiration however did not yield conclusive pictures. Hepatitis, cytolytic or mixed type, was quasi constant (96%), although clinically symptomatic in only 6 cases, and classified as severe (grade 4) in 3. Elevated alcaline phosphatase and gamma glutamyl transpeptidase were remarkably persistent in some cases (> 6 months in four cases, and up to one year in one). No liver biopsy was performed. Moderate renal failure was less frequent (46%), but notably present in 5 of the 7 allopurinol‐induced cases, serum creatinine ranging from 140 to 270µmol\l. Eosinophilia was found between 500 to 1500\mm3 in 5 cases and was above 1500 mm³ in 22 cases.

The course of DHS was remarkably prolonged, complete clinical recovery taking from 14 to 120 days (mean 66 ± 41 days). More chronic cases (more than 6 to 8 weeks) were no different from the other cases in terms of drug intake, eosinophilia, age, and treatment with corticosteroids. Two patients (aged 57 and 72 years), both with grade 4 hepatitis and consecutive coagulation disorders, died respectively from septic shock and cerebral haemorrhage. Nineteen patients (67%) were given prednisone at a dose of 0.5‐1 mg\kg\d, for a duration of 10 to 120 days. Steroid prescription was empirical, based on merely subjective criteria including: impression of severity, worsening picture, patient‘s request for more aggressive treatment, and finally patient‘s and\or physician‘s impatience. In most cases, steroids were rapidly effective (within 1‐4 days) on the main clinical symptoms i.e. fever, rash and malaise, but not on the liver enzymes. However, they did not show any consistent preventive effect: 5 patients developed the syndrome, and two others died while under steroids. In addition, rebound phenomena occurred in 6 cases (one fatal) when steroid dosages were tapered.

Table II details the causative drugs.Table II. Incriminated Drugs and Appropriateness of Prescriptions

Incriminated Drugs	Number	Inappropriate indication
Carbamazepine	8	4\8
Allopurinol	7	6\7
Minocycline	4	4\4
Dapsone	2	0\2
Salazopyrine	2	2\2
Other (one case each): Aspirin\Barbiturates\Nevirapine\Sulfamethoxazole‐Trimethoprime	4	2\4
Undetermined	1
Total	28	18\28 (64%)

.

Three drugs, namely carbamazepine, allopurinol, and minocycline accounted for 2\3 of all DHS cases. Ten patients (mean age 31) had taken the incriminated drug in a single drug regimen, whereas 18 patients (64%), mostly older ones (mean age 57.5) took it in multi‐drug associations (an average of 4 drugs\patient, and up to 10 drugs). The latter included all the most widely prescribed ones: analgesics and aspirin, anti‐infective, anti‐hypertensive, diuretics, antihistamines and steroids, benzodiazepines and neuroleptics among others. With one exception (one fatal case), chronological as well as evidence‐based arguments were concordant enough to incriminate one specific drug. Of interest, three incidental drug re‐exposures occurred that resulted in prompt relapse, with incubation times of 2‐9 days, that appeared strikingly shortened as compared to primary exposure. Finally, drug prescription was judged inappropriate in 64% of the cases (Table II). Most frequent inappropriate indications included: carbamazepine in non neuropathic pain or depression; minocycline in common acne, or non bacterial prostatitis; allopurinol in acute gout attack or in symptomless hyper‐uricemia; and finally, salazopyrine in ill defined rheumatic pain.

Discussion

We will briefly comment on epidemiology and ethnic distribution of the DHS, its relation to SLE, management, and finally on the incriminated drugs and prevention. Because a single organ dysfunction can be most apparent, DHS may be misclassified under vague categories such as "allergic reaction", "drug rash", or "drug related hepatitis", and may therefore be underestimated. Occurrence rates have been documented for some specific drugs, roughly between 1\1,000 and 1\10,000 exposures [10], notably for anticonvulsivants (0.2‐1\1000) [1, 17], minocycline (0.52\1000) [8], dapsone (5‐10\1000 prescriptions) [11], and lately abacavir (peaking at 30 to 50\1000) [9, 18]. Global incidence including all drugs, on a general population base, has not been previously assessed. The present 0.9\100,000 estimation is higher than expected for a syndrome considered as rare. It may represent a minimal estimation due to the restrictive inclusion criteria, and the exclusion of outpatients. In our study, DHS appeared as the most frequent systemic drug reaction to occur in West‐Indians. Afro‐Caribbeans tended to be over‐represented (75% of the population, 93% of the cases) as compared to non Afro‐Caribbeans. Due to the small sample size, the difference, however, was not statistically significant. Anyhow, interpreting ethnic data always deserves particular caution, due to the scientifically irrelevant concept of race, notably in populations with large racial intermixing such as the West‐Indian one [19]. However, this point raises interesting questions about the genetic basis for drug reactions: enhanced prevalences have effectively been documented between twins [20], within families [17], or in some ethnic groups. As examples, the dapsone syndrome has been documented in Melanesian leprosy patients with a 24 fold increased incidence as compared to the average in other settings [11], and also acute angio‐oedema induced by angiotensin converting enzyme inhibitors that seems to be more prevalent in Blacks [21]. The latter have also been considered more susceptible to phenytoin or minocycline DHS [1, 3, 12, 13], although haplotypes responsible for this susceptibility have not been identified. Recent advances in the pathophysiology of the DHS have indeed focused on processes that are strongly affected by genetic variability: mainly those involving the immune system [3, 4, 12] and the hepatic microsomial cytochrome P450 enzymatic system [9, 22, 23]. Finally, a unique association with strong predictive values has been recently established between hypersensitivity syndrome (to abacavir) and several MHC haplotypes, namely: HLA‐B57O1, HLA‐DR7, and HLA‐DQ3 [24].

Another finding of this study was persistent antinuclear antibodies ‐‐ but not anti‐DNA ‐‐ which were found in 5 patients, 2 of them previously known to have systemic lupus erythematosus (SLE). This may suggest either a possible overlap of DHS and SLE, particlarly drug induced‐LE, or a particular susceptibility of lupus patients to DHS. Both conditions indeed, may share some clinical features as well as some inducing drugs, notably minocycline and allopurinol [25‐28], and finally, DHS has regularly been described as a "lupus‐like syndrome" before being properly identified as a distinct entity [8]. Moreover, SLE is also known to have higher prevalence and morbidity rates in subjects of African descent [29].

Our experience in steroid management of DHS, leads us to emphasize that although of common use, and of apparent usefulness, steroid therapy has not so far been supported by randomised studies, and is still empirical and questionable [1‐5, 13, 17, 30].

With regard to causative drugs, our study confirms the rank of minocycline, which is neither a new nor a restricted drug, but has only lately been identified as a major cause of DHS [13, 29]. By contrast, nevirapine and abacavir, two different recently licensed antiretrovirals, have quickly appeared as the drugs yielding the highest rates of DHS ever reported (around 3‐5%), a point that resulted in prompt and appropriate recommendations [9, 18, 24, 31]. Of final concern in the epidemiology of drug reactions, is the major fact of inappropriate drug prescription [22]. Our results suggest that, in theory, about two thirds of DHS cases could be avoided by more rational and more cautious prescriptions. Carbamazepine, allopurinol, and minocycline for instance, should no longer be used as first‐line drugs, because alternatives are available in most clinical situations [17, 26, 32, 33].

In summary, DHS may exhibit an unexpectedly high incidence in Afro‐Caribbean populations, in which it appeared as the most frequent type of severe drug reaction (roughly twice as frequent as Lyell‘s syndrome). The present study however, could not definitively confirm the alleged over‐susceptibility in Blacks, a point that deserves further and larger comparative studies. The trio: carbamazepine‐allopurinol‐minocycline, as is the case elsewere, accounted for two thirds of the cases. The conclusion is that DHS is largely avoidable, just by following the "good prescribing" principles [34, 35] that can be summarized as follows: few drugs; only for approved indications; careful assessment of tolerance and of risk\benefit ratio; high suspicion of drug reaction in any ill‐explained clinical event; and finally, definitive withdrawal of any suspect drug.

References

1 . Roujeau JC, Stern RS. Severe adverse cutaneous reactions to drugs. N Engl J Med 1994; 331: 1272‐85.

2 . Sontheimer R, Houpt KR. DIDMOHS: A proposed consensus nomenclature for the Drug Induced Delayed Multiorgan Hypersensitivity Syndrome. Arch Dermatol 1998; 134: 874‐6.

3 . Bocquet H, Bagot M. Drug‐induced pseudolymphoma and hypersensitivity syndrome (drug rash with eosinophilia and systemic symptoms: DRESS). Seminars Cutaneous Medicine and Surgery 1996; 15: 250‐7.

4 . Callot V, Roujeau JC, Bagot M, Wechsler J, Chosidow O, Souteyrand P, Morel P, Dubertret L, Avril MF, Revuz J. Drug induced pseudo‐lymphoma and hypersensitivity syndrome. Arch Dermatol 1996; 132: 1315‐21.

5 . Vittorio CC, Muglia JJ. Anticonvulsivant hypersensitivity syndrome. Arch Intern Med 1995; 155: 2285‐90.

6 . Prussik R, Shear NH. Dapsone hypersensitivity syndrome. J Am Acad Dermatol 1996; 35: 346‐9.

7 . Arellano F, Sacristan JA. Allopurinol hypersensitivity syndrome: a review. Ann Pharmacoth 1993; 27: 337‐43.

8 . Sturkenboom MCJM, Meier CM, Jick H, Stricker BHC. Minocycline and lupus‐like syndrome in acne patients. Arch Intern Med 1999; 159: 493‐7.

9 . Carr A, Cooper DA. Adverse reactions to antiretroviral therapy. Lancet 2000; 356: 1423‐30.

10 . Knowles SR, Uetrecht J, Shear NH. Idiosyncratic drug reactions: the reactive metabolite syndromes. Lancet 2000; 356: 1587‐91.

11 . Reeve PA, Ala J, Hall JJ. Dapsone syndrome in Vanuatu: a high incidence in multiple drug treatment of leprosy. J Trop Med Hyg 1992; 95: 266‐70.

12 . Silverman AK, Fairley J, Wong RC. Cutaneous and immunologic reactions to phenytoin. J Am Acad Dermatol 1998; 18: 721‐41.

13 . Beneton N, Bocquet H, Cosnes A, Revuz J, Roujeau JC. Syndrome d‘hypersensibilité à la minocycline: 7 cas. Ann Dermatol Venereol 1997; 124: 69‐70.

14 . Begaud B, Evreux JC, Jouglard J, Lagier G. Imputabilité des effets inattendus ou toxiques des médicaments. Therapie 1985; 40: 111‐8.

15 . Moore N, Paux G, Begaud B, Biour M, Loupi E, Boismare F. Adverse drug reaction monitoring: doing it the French way. Lancet 1985; II: 1056‐8.

16 . VIDAL, Dictionnaire, 74è édition, OVP‐Edition du Vidal, Paris 1999.

17 . Knowles SR, Shapiro LE, Shear NH. Anticonvulsivant hypersensitivity syndrome: incidence, prevention and management. Drug Saf 1999; 21: 489‐501.

18 . Escault L, Liotier JY, Albengrezs E, Cheminot N, Vittecoq D. Abacavir rechallenge has to be avoided in case of hypersensitivity reaction. AIDS 1999; 13: 1419‐20.

19 . Senior PA. Ethnicity as a variable in epidemiological research. BMJ 1994; 309: 327‐30.

20 . Edwards SG, Hubbard V, Aylett S, Wren D. Concordance of primary generalised epilepsy and carbamazepine hypersensitivity in monozygotic twins. Postgrad Med J 1999; 75: 680‐1.

21 . Gibbs CR, Lip GY, Beevers DG. Angio‐oedema due to ACE inhibitors: increased risk in patients of African origin. Brit J Clin Pharmacol 1999; 48: 861‐5.

22 . Pirmohamed M, Breckenridgd AM, Kitteringham NR, Park K. Adverse drug reactions. BMJ 1998; 316: 1295‐8.

23 . Pirmohamed M, Madden S, Park BK. Idiosyncratic drug reaction: metabolic bioactivation as a pathogenic mechanism. Clin Pharmacokinet 1996; 31: 216‐30.

24 . Mallal S, Nolan D, Witt C, Masel G, Martin Am et al. Association between presence of HLA‐B5701, HLA‐DR7, and HLA‐DQ3 and hypersensitivity to HIV‐1 reverse transcriptase inhibitor abacavir. Lancet 2002; 359: 72‐32.

25 . De Vriese SP, Philippe J, Van Renterghem DM, De Cuyper CA, Hindryckx PH, Mathys EG, Louagie A. Carbamazepine hypersensitivity syndrome: report of 4 cases and review of the literature. Medicine 1995; 74: 144‐51.

26 . Byrne PAC, Williams BD, Pritchard MH. Minocycline related Lupus. Brit J Rheumatol 1994; 33: 674‐6.

27 . Hess E. Drug‐related lupus. N Engl J Med 1988; 318: 1460‐2.

28 . Ward MM, Studenski S. Clinical manifestations of systemic lupus erythematosus. Identification of racial and socio‐economic influences. Arch Intern Med 1990; 150: 849‐53.

29 . Elkayam O, Levartovski,D, Yaron M, Burke M, Vardinon M, Capsi D. Clinical and immunological study of 7 patients with minocycline‐induced auto‐immune phenomena. Am J Med 1998; 105: 484‐7.

30 . Roujeau JC. Treatment of severe drug eruptions. J Dermatol 1999; 26: 718‐22.

31 . BourezaneY, Salard D, Hoen B, Vandel S, Drobachef C, Laurent R. DRESS syndrome associated with nevirapine therapy. CID 1998; 27: 1321‐2.

32 . Handke KR, Noone RM, Stone WJ. Severe allopurinol toxicity: description and guidelines for prevention in patients with renal insufficiency. Am J Med 1984; 76: 47‐56.

33 . Gottlieb A. Safety of minocycline for acne. Lancet 1997; 349: 374.

34 . Buetow SA, Sibbald B, Cantrill JA, Halliwell S. Prevalence of inappropriate long term prescribing in general practice in the United Kingdom. BMJ 1996; 313: 171‐4.

35 . Barber N. What constitutes good prescribing ¿ BMJ 1995; 310: 923‐5.